A Kimmeridgian time-slice through the Humber Group of the central North Sea: a test of sequence stratigraphic methods

Abstract

Abstract Reservoir quality of the Fulmar Formation in the UK Central Graben depends on the original distribution of facies as well as subsequent modification by partly facies-dependant diagenesis. Numerous sequence stratigraphical analyses of the Humber Group have been published recently, each of which has elements which are convincing geologically. Nevertheless, it is our experience that facies and reservoir prediction within the Humber Group remain difficult. Sequence stratigraphical interpretations in the Group are clearly non-unique due to major uncertainties in the data. (1) Biostratigraphical uncertainties include species concepts, recognition of agediagnostic taxa in thermally altered floras, validity of repeated micropaleontological events and equivocal paleoenvironmental information. (2) Sedimentological uncertainties are due to the gross facies similarity of the Fulmar Formation over a large geographic area and throughout the Upper Jurassic, which make the detection of potential maximum flooding surfaces and especially sequence boundaries difficult even in many cored wells. Furthermore strong syndepositional subsidence variations due to both salt movement and tectonics, lead to aliased sampling of the subsurface in wells, which are mostly situated on highs. (3) Seismic correlation difficulties are due to persistent seismic imaging problems below the Late Cimmerian (‘X’) Unconformity. (4) The depositional and sequence stratigraphical models for Humber Group deposition need to be set in the context of a transgressively skewed, back-stepping system with strong local subsidence control. The overall transgression is thought to be due to the combination of sea-level rise, widening and deepening of the basin due to tectonics, and reduced sediment supply through time. In this situation existing models have to be applied with even more care than normal if they are not to be misleading (e.g. concepts of forced regressions). These difficulties are clarified in the context of a Kimmeridgian time slice through the Humber Group. Multiple options exist in well log correlations. As a result palaeogeographic reconstructions of the time-slice have an error margin of ±15 km, which has serious implications on the validity of reservoir quality predictions.